Cyclic nucleotide phosphodiesterases (PDEs) are the key enzymes that control the cellular concentration of "second messengers" adenosine or guanosine 3', 5'-cyclic monophosphate (cAMP or cGMP). The human genome encodes 21 PDE genes and over 60 PDE isoforms categorized into 11 families. All PDEs contain a conserved catalytic domain, but each family possesses individual substrate specificity and selective inhibitors. Selective inhibitors of PDEs have been widely studied as therapeutic agents for various diseases. For example, PDEs inhibitor sildenafil (VIAGRA(tm)) is a drug for erectile dysfunction and PDE3 inhibitor cilostazole (Pletal(tm)) is a drug for intermittent claudication. The wide medical applications of PDE inhibitors have attracted great attention from both academic and industrial research groups. However, it has been mysteries how the similar active sites of PDEs distinguish the different substrates and inhibitors. We hypothesize that the substrate specificity and inhibitor selectivity are determined by both the chemical nature of active site residues and the conformations of the PDE active sites. This proposal chooses cAMP specific PDE4 and cGMP specific PDE5 and PDE9 as the target systems to study the substrate specificity and inhibitor selectivity with approaches of crystallography and protein engineering. The structures of PDE4, PDES and PDE9 in complex with substrate, substrate analogues, and selective inhibitors will be determined. The candidate residues will be switched between PDE4 and PDE5 by a single or multiple mutations for further illustration of the substrate specificity. The structures in this proposal, together with those from the last funding period, will reveal key residues and elements for determination of the substrate specificity and identify potential subpockets contributing to the selective binding of the inhibitors. Since the inhibitor selectivity is a key issue for side effects of drugs, the structures of PDEs in complex with inhibitors will provide templates for design of family- or subfamily-selective inhibitors and ultimately improve the drugs efficiency for treatment of the diseases.